Re: [commit: ghc] master: define own version of PRIdPTR on platform where its not available (71f7ab6)
Gabriel Dos Reis <[email protected]>
| Newsgroups | gmane.comp.lang.haskell.cvs.ghc |
|---|---|
| Message-ID | <CAAiZkiBwzYwSExSaCusvpVozmnntxGCPg_OoYSoFE2zGxcuNHw@mail.gmail.com> |
On Fri, Nov 9, 2012 at 1:54 PM, Karel Gardas <[email protected]> wrote: > On 11/ 9/12 08:33 PM, Gabriel Dos Reis wrote: >> >> On Fri, Nov 9, 2012 at 12:18 PM, Karel Gardas<[email protected]> >> wrote: >>> >>> On 11/ 9/12 06:53 PM, Gabriel Dos Reis wrote: >>>> >>>> >>>> Include<inttypes.h> before all other standard headers. >>>> >>> >>> I I understand your "standard headers" well, then this is already done in >>> mkDerivedConstants.c, see: >>> >>> [...] >>> #define PROFILING >>> #define THREADED_RTS >>> >>> #include "PosixSource.h" >>> #include "Rts.h" >>> #include "Stable.h" >>> #include "Capability.h" >>> >>> #include<inttypes.h> >>> #include<stdio.h> >>> #include<string.h> >>> [...] >>> >>> The problem which I describe is in the position of PosixSource.h which >>> defines problematic _XOPEN_SOURCE and<inttypes.h> includes... >>> >>> Karel >> >> >> The point I was making (admittedly imperfectly) is >> that<inttypes.h> should be included before "PosixSource.h". >> Does that work for you? > > > This works well, but emits warning: > > In file included from includes/ghcconfig.h:4:0: > 0, > from includes/Stg.h:60, > from includes/Rts.h:30, > from includes/mkDerivedConstants.c:27: > > includes/ghcautoconf.h:388:0: > warning: "_FILE_OFFSET_BITS" redefined > > /usr/include/sys/feature_tests.h:204:0: > note: this is the location of the previous definition > > Karel aha, so this actually uncovered a potential problem -- a problem that is benign in this case because I think the redefinition is to the same value. Why is ghcautoconf.h defining _FILE_OFFSET_BITS? It looks like something that should come from system headers. -- Gaby